Changes in motor and cognitive functions that occur with the aging process negatively affect individuals' mobility and balance performance, there by increasing the risk of falls. Today, it is widely accepted that gait and postural control are not merely motor processes, but also require the active participation of attention and executive cognitive processes. In this context, the dual-task approach-wherein an individual simultaneously performs a second cognitive or motor task during gait-has become a widely utilized method for evaluating cognitive-motor interaction in older adults (Muir-Hunter \& Wittwer, 2016). Performance degradation under dual-task conditions is explained by the individual having to allocate limited cognitive resources between two tasks, a phenomenon that becomes particularly more pronounced in older adults. Studies in the literature examining the relationship between dual-task performance and falls demonstrate that performance changes occurring under dual-task conditions are more strongly associated with falls compared to single-task conditions (Muir-Hunter \& Wittwer, 2016). However, there is still no full consensus on which protocol, which secondary task, and which performance outcome are clinically most meaningful in dual-task assessments. The dual-task Timed Up and Go (TUG) test stands out as a practical and clinically applicable method for evaluating functional mobility. Tang et al. (2014) showed that motor dual-task TUG performance was superior to single-task TUG in discriminating the state of prefrailty in middle-aged and older adults. In the same study, it was reported that the ROC analysis of dual-task TUG performance showed significant discriminative power, and slow performance significantly increased the likelihood of prefrailty. These findings reveal that dual-task-based measurements are not only explanatory but also possess the potential to serve as classification tools and contribute to clinical decision-making processes. The relationship between cognitive functions and dual-task performance is another notable topic in the literature. It has been shown that both gait performance and cognitive performance deteriorate more markedly under dual-task conditions in individuals with mild cognitive impairment (Muir-Hunter \& Wittwer, 2016). Furthermore, it is indicated that the type and difficulty level of the secondary task used have significant effects on dual-task performance. This situation highlights the critical importance of standardizing dual-task assessments. Fear of falling is also a major factor affecting functional performance in older adults. In a study conducted by Sapmaz et al. (2021) on institutionalized older adults, it was shown that individuals with a fear of falling had worse single-task and dual-task TUG performances; additionally, their balance and mobility levels were also lower. These findings demonstrate that the fear of falling is not merely a psychological condition but is closely related to functional capacity. However, a large part of current studies remains at the group-comparison level, and analyses modeling the role of dual-task performance in predicting clinical fall risk remain limited. Recent studies show that dual-task assessments are highly sensitive in uncovering hidden functional impairments in older adults (Falbo et al., 2016). Especially in frail populations such as institutionalized older adults, dual-task performance is thought to be closely associated with falls. However, most of these studies either feature intervention-based designs or are limited strictly to correlation analyses. A prominent gap in the literature is the scarcity of studies evaluating fall risk via dual-task cost (which represents the relative change compared to single-task performance) rather than absolute measurements of dual-task performance. Because dual-task cost reflects the relative deterioration in an individual's performance under cognitive load, it may serve as a clinically more meaningful indicator. Nevertheless, studies examining the role of dual-task cost in discriminating fall risk and establishing a cross-sectional classification model based on this variable remain highly limited.
